Lake Processes Simulations with the Weather Research Forecasting Model for the Great Lakes

Post on 24-Feb-2016

31 views 0 download

Tags:

description

Lake Processes Simulations with the Weather Research Forecasting Model for the Great Lakes Jiming Jin and Hongping Gu Departments of Watershed Sciences and Plants, Soils, and Climate Utah State University Yhua Wu and Michael B. Ek - PowerPoint PPT Presentation

transcript

Lake Processes Simulations with the Weather Research

Forecasting Model for the Great Lakes

Jiming Jin and Hongping Gu Departments of Watershed Sciences and Plants, Soils, and Climate

Utah State University

Yhua Wu and Michael B. EkEnvironmental Modeling Center, National Centers for Environmental

Predictions

• Bias in reanalyzed lake surface temperature

• Coupling of WRF and a lake model

• Calibration of the coupled WRF-lake model

• Lake-effect precipitation simulations

• Summary

Outline

Quantifying lake processes and their effects on local and regional weather and climate using the Weather Research and Forecasting (WRF) model coupled with a physically-based lake model.

Lake-Effect Snow

Objective

Difference of the NARR LST and MODIS LSTLake region mean

Superior

Huron Ontario

ErieMichigan

LST: Lake surface temperature

NARR: North American Regional Analysis

MODIS: Moderate Resolution Imaging Spectroradiometer Satellite data

OBS- U. Delaware 87mm

WRF-MODIS89mm

WRF-NARR95mm

Precipitation Simulations at 10 km resolution with WRF over the Great Lakes Region

February 2006

All the release versions of the WRF model do not include a lake scheme.

The lake surface temperature is provided by the forcing data for the WRF model.

Regional Climate Model

• We have coupled a lake model into WRF 3.2

• This lake model is a one-dimensional water and energy balance model (Hostetler et al., 1993; 1994).

• The lake in the model is divided into 10 vertical layers.

• Ice fraction and snow on the ice are also considered in the lake model.

A Physically-based Lake Model

Surface Temperature Simulations at 10 km resolution for Lake Erie

The average depth of Lake Erie is 19 m with a maximum depth of 64 m

Surface Temperature Simulations at 10 km Resolution for Lake Superior

The average depth of Lake Superior is 147 m with a maximum depth of 406 m.

Processes Affecting Lake Surface Temperature

Atmospheric Forcing Atmospheric Forcing

Shallow Lake

Deep Lake Deep Water Forcing

Calibration of Lake Physical Parameters

Old New

Eddy Diffusivity Too small Enlarged

Roughness length 1 cm 0.5 cm

Bathymetry 50 m Actual data

Surface Temperature Simulations for Lake Superior

Near-Surface Air Temperature Simulations for Lake Superior

The air temperature was measured or simulated at a height of 4 m

NARR minus MODIS WRF-Lake minus MODIS

Winter (DJF), 2003-2008 Winter (DJF), 2003-2008

Mean Bias = 5.2 oC Mean Bias = 2.2 oC

Lake Surface Temperature Bias

Dec Dec

Jan Jan

Feb Feb

Observation Simulation

Dec

Jan

Feb

2003-2008

Lake Ice Fraction Simulations

0.55

0.82

0.82

mm

Dec Jan Feb

OBS

WRF-Lake

WRF-NARR

2003-2008

Winter Precipitation Simulations

Precipitation Bias (Domain Average)

WRF-NARR minus OBS

WRF-Lake minus OBS

2003-2007

Bias = 7.2mm/month

Bias = 4.3mm/month

Simulated Precipitation Bias

The coupled WRF-Lake model realistically simulates the lake surface temperature and lake ice fraction for the Great Lakes.

This coupled model also reduces the biases in the lake-effect precipitation simulations.

Summary

Acknowledgement

This project is supported by the NOAA Modeling, Analysis, Predictions, and Projections Program. Grant Number: NA090AR4310195.